Please create an account to participate in the Slashdot moderation system

 



Forgot your password?
typodupeerror
Communications Science Technology

Government Funded Atomic Clock On a Chip 134

Posted by samzenpus
from the let's-get-small dept.
An anonymous reader writes "Today most applications that require accurate atomic clock readings — from sorting separately routed telecommunications packets to timing simultaneous demolition charges — usually refer to signals from global positioning systems (GPS). For applications where GPS is unavailable, such as indoors, underground, undersea or on the battlefield where electronic jamming is present, large, heavy, power hungry hardware atomic clocks were needed. Now an atomic clock-on-a-chip is available that is the result of 10 years of government-funded research and development. The chip is not cheap — $1,500 — but it costs less than conventional atomic clocks and the price is sure to go down as manufacturing gears up to meet demand from military applications."
This discussion has been archived. No new comments can be posted.

Government Funded Atomic Clock On a Chip

Comments Filter:
  • The original press release [symmetricom.com] is from January 18th 2011. Just sayin'. Of course this is a very nifty device and all that.

  • For those of us who need accurate clocks and don't have $1500 to spend, highly stable temp controlled oscillator chips are cheap and common right now. (Search eBay for OXCO)

    For example, this one [ebay.com] (which I'm using) is accurate in the PPB range:

    • by imlepid (214300)

      Yes, you can use OCXOs, but they aren't technically atomic clocks. Further, an OCXO (like the one you showed) requires 1.5W, which doesn't sound like much, but the unit linked to above needs only 100mW. A true atomic clock (a rubidium oscillator [ebay.com], for example) is significantly larger than this unit and also draws much more power (11W, steady state).

      All things told, though, a OCXO or rubidium frequency standard from eBay should be good enough for most users.

    • by Chris Burke (6130)

      Well, that claims 20 ppb, so that's ~ 1 part in 10^8. If the article mentioned the stability of this chip, I missed it, but other cesium atomic clocks are stable to 1 part in 10^14. So they're literally orders of magnitude more precise.

      But if you 10^8 is good enough, then $20 sounds like a great deal!

      I'm figuring they wanted to develop this chip for applications where currently caesium or other atomic clocks are required, though.

    • by kaiser423 (828989)
      The chip from the article is ~100x more accurate and consumes 1/10th the power and is similar size of smaller for ~75x the price.

      Pretty big leap if you ask me, and I know lots of people that will be looking to utilize these. I expect in the near future that the pricing will drop significantly, since that pricing was just for the initial batch run.
      • by blair1q (305137)

        I know lots of people that will be looking to utilize these.

        For what?

        I don't know of much that needs timing precision better than a few ppb. If it does, it probably needs a redesign. Because even with a few ppb, the chances the clock is simply set wrong will be higher than the chance the drift will take it out of the valid range, and being robust enough to handle either case is probably worth the $1500.

        • by Arlet (29997)

          I guess that most people requiring these clocks are not really interesting in having a correct absolute time, but rather a very low drift to allow measuring intervals at really high precision.

          And if absolute time matters, you can always calibrate it with another atomic clock, or GPS signal.

        • by mellon (7048)

          They're really handy for detecting time dilation caused by variations in the gravitational field. It wouldn't surprise me if, when the price comes down, people start using them to survey construction sites for geological stability. Try doing that with some wimpy OXCO that's only accurate to a few parts per billion.

          • by blair1q (305137)

            They're going to need a thermometer that goes to parts per trillion, then, too, because the Earth reacts to the heat of the sun in a way that looks just like low-level geological instability.

            As for time dilation of any kind, there are maybe five people (all of them Doctors of some sort) who care in anything other than a merit-badge sort of way.

            • As for time dilation of any kind, there are maybe five people (all of them Doctors of some sort) who care in anything other than a merit-badge sort of way.

              Who's one of them?

              • Who's one of them?

                Yes, that's right, Dr. Who.

              • The people who design GPS satellites for a start. And I imagine other satellites requiring precise timing information out beyond the Earth's gravitational field.

              • by Twisted64 (837490)

                As for time dilation of any kind, there are maybe five people (all of them Doctors of some sort) who care in anything other than a merit-badge sort of way.

                Who's one of them?

                Who's all of them!

              • by Anonymous Coward

                BZZZZZZZZZZZZT!

                Thank you for playing!

                You're wrong, though.

                Time Dilation must be accounted for in GPS systems! If you use GPS, you're using systems that need to account for time dilation. Here's an explanation

                Because of relativistic effects, clocks on-board each GPS satellite should tick faster than identical clocks on the ground by about 38 microseconds per day (45-7=38).

                This sounds small, but the high-precision required of the GPS system requires nanosecond accuracy, and 38 microseconds is 38,000 nanosecon

              • And someone moonlighting as an English teacher [youtube.com], presumably.
            • The on-board GPS clocks are corrected for time dilation due to their orbital velocity. By extension, everyone who uses GPS benefits from and has an interest in that being accounted for.

          • by careysub (976506)

            They're really handy for detecting time dilation caused by variations in the gravitational field....

            And the dilation caused by motion. Note that one of the organizations developing this is Draper Laboratories - home of the world's best inertial navigation systems. This would be a crucial component for a new compact, low cost (as military equipment goes) ultra-accurate inertial guidance system for weapons that is NOT dependent on GPS. Ever heard of GPS jammers?

        • by Agent0013 (828350)

          For what?

          If you can put an atomic clock into a GPS unit, you can match the timing signal on more than just the number pattern. You can match on the square wave train. This allows you to get into millimeter accuracy. For use in car navigation, not neccessary. But if you want to do survey work, it could be very usefull.

      • by tibit (1762298)

        Inertial navigation needs that sort of accuracy. Low-drift, high-resolution laser gyros are IIRC as good as and no better than their time bases are. Drift in reference frequency causes changes in gain...

      • Yeah, the next iPhone will include one, and prices will plummet.
    • For those of us who need accurate clocks and don't have $1500 to spend, highly stable temp controlled oscillator chips are cheap and common right now. (Search eBay for OXCO)

      For example, this one [ebay.com] (which I'm using) is accurate in the PPB range:

      An OXCO has excellent phase noise and short term stability, but drifts. A Cs reference has better long term stability. An Rb device has pretty good long term stability once you factor out the linear "aging" (i.e., a constant change). GPS-DO combine GPS receivers with firmware and discipline another oscillator, usually a quartz crystal because of the low phase noise.

      So the right way to compare these various references is with an Allan deviation plot, which rates the stability across different time scales.

  • by LWATCDR (28044) on Wednesday May 11, 2011 @01:35PM (#36097198) Homepage Journal

    I can think of two to uses off the top of my head. The first is for really fast frequency hopping radios. The rate at which they can hop from one to the next has got to be in some measure limited to how accurate the clock they use is.
    And the next one would be improved navigation. You could use these with ground stations and provide extremely accurate navigation and you could use more powerful transmitters so they would be harder to jam.
    Now if they could uses these to put a time signature on every radio, tv, and cell tower You could improve navigation in areas where GPS doesn't work so well. Like in buildings. cites with lots of tall buildings, or areas with lots of tree cover.

    • i wonder (and I am NOT saying anything educated but I wonder) if you could simplify the GPS signal (which currently has to send time, and that is used to get distance) if you had an accurate clock in the receiver. If you could simplify the signal, could you make GPS more accurate or have much better reach in terms of reception?

      i'm sure there are tons of uses for a good atomic clock, but this leapt to mind. If someone can say why i'm wrong, it would teach me much :)
      • by Arlet (29997) on Wednesday May 11, 2011 @02:05PM (#36097622)

        Having a local known good time would reduce the GPS error by itself. It would also allow 3D position to be determined with 3 visible satellites instead of 4.

        • That makes sense! I appreciate slashdot for these little things. I have read about GPS, but didn't derive these two characteristics. I remember something about quantum entanglement one day allowing for a more sensitive read of when the signal arrived, I have read about WAAS.. and when the prices of atomic clock chips come down I hope they are used in just this way.

          //Satnav has always fascinated me. I use a Miomap-Digiwalker flashed with a WinCE shell featuring IGO8 (best nav soft on standalone GPS's in m
        • by Andy Dodd (701) <atd7&cornell,edu> on Wednesday May 11, 2011 @02:59PM (#36098310) Homepage

          Yup. One of the barriers to postprocessing is local clock inaccuracy - so having a local atomic clock would be great for survey-grade GPS units.

          And as you stated - if receiver clock offset is 0, then you don't need to solve for it, and can get 3D position with 3 sats instead of 4. The actual effect of an inaccurate clock on the error is harder to determine - I have a feeling that with a reasonable quality local crystal oscillator (good enough not to cause cycle slips in the measured carrier phase, etc.) it's insignificant compared to ionospheric error and RF noise in the pseudoranges, along with multipath. The new L2C civilian signal will help some of these issues.

          A highly accurate local clock might also make dead reckoning in a blockage situation (urban canyons, tunnels, etc) and signal reacquisition after blockage goes away faster.

        • If I understand the math right, it'd also let you almost determine 2D position with two rather than three, which means faster locks. I say almost, because the equasions would actually provide two solutions - but you can handle that in software, by using the last-known-good measurement to determine which of those two is correct.
      • by woolpert (1442969)

        Except this isn't a problem that needs solved.

        Kinematic GPS solves position using the carrier instead of the (time) code. Since the carrier length is ~19cm you're instantly much tighter than is possible solving with the time code.

        And, as others above me have said, (regardless if you're carrier or code locked) a 3D solution needs four birds visible because of the receiver's inaccurate clock. At best an atomic clock on the receiver means one less variable to solve for = one less bird needed. Today when we

    • by blair1q (305137)

      At some point "really fast frequency hopping" becomes its own modulation trope. I.e., you aren't using the frequencies you're hopping on as the carrier, you're using the pattern of your hopping. So maybe. And I call dibs on the prior art if frequency-hopping-modulation works. We just can't call it FHM or nobody will get any work done when googling for the literature.

      DGPS takes the slop out of GPSS, and I'm not sure if one of these chips could possibly improve on it, though maybe, by providing a base ref

      • by LWATCDR (28044)

        GPS uses a signal that has poor penetration and is also very weak. TV, Radio, and Cell signals are lower frequency and penetrate buildings better.

    • by mortonda (5175)

      Not being very familiar with relativity and all that, at what precision and time frame might relativistic effects cause this to become unsynchronized?

      • Not being very familiar with relativity and all that, at what precision and time frame might relativistic effects cause this to become unsynchronized?

        Hmm, rough estimate says 10E-14 scale error will be detectable at >150 km/hour.

      • by Coren22 (1625475)

        unsynchronized in relativity to what? GPS sats are moving pretty fast and don't get unsynced.

      • by Arlet (29997)

        As long as you know your speed, it's not a problem to compensate for relativistic effects. GPS systems must already do that anyway.

    • by Agripa (139780)

      I can think of two to uses off the top of my head. The first is for really fast frequency hopping radios. The rate at which they can hop from one to the next has got to be in some measure limited to how accurate the clock they use is.

      Frequency hopping spread spectrum, direct sequence spread spectrum, and TDMA systems can phase lock their local oscillators together so absolute accuracy is not required. Performance is only limited by local oscillator phase noise. An atomic oscillator would help with synchro

  • "The chip is not cheap--$1500--but it costs less than conventional atomic clocks and the price is sure to go down as manufacturing gears up to meet demand from military applications start using it."

    For the price to come down we'll have to wait for the Chinese to finish tooling their new plant.

    • by blair1q (305137)

      How long do we have to wait?

      Please express your answer to 15 significant figures.

  • No! not Symmetricon! (Score:4, Interesting)

    by NixieBunny (859050) on Wednesday May 11, 2011 @01:55PM (#36097464) Homepage
    Symmetricon has been buying up all the other precision clock makers, and is now a monopoly. They can and do charge whatever they like for such products.
    • Symmetricon has been buying up all the other precision clock makers, and is now a monopoly. They can and do charge whatever they like for such products.

      Goddamnit. Do you have to go do this? Here we have an interesting, techy article and you go rain on everybody's parade by trotting out some little factoid that either makes the tech responsible for the imminent demise of the planet or at the very least points out some totally unsavory bit about the company that manufacturers it. Can't you just leave it at Apple, Microsoft and Google?

      Give us a break for just a bit, will you?

      • I'd love to give it a break, but every time I turn around they've ruined another perfectly good product designed by someone else.
      • by satan666 (398241)

        I know what you're saying and all but this is bigger that you initially thought.
        Let's dig a bit deeper.

        When GPS first came out (and President Clinton allowed for the superior GPS signal to be released to the public), we all thought: Wow! Finally!
        As EE's we thought the time had come for really, really cheap GPS based NTP servers and timing products. You know, like in the $100.00 range.

        Who gives a fuck? Well, actually we all do. We are all paying for NTP synchronization hardware because all major server farms

  • This guy has had a real atomic wristwatch for at least a decade. http://leapsecond.com/pages/atomic-bill/ [leapsecond.com]
  • It's an ATOM processor.

  • Seriously, this is much cheaper than some of the Wrist borne Chronographs... Tag Heuer look out!

    Seriously though, I mean, sure you'll need a few more electronics and such to get it to show time, but over all, it wouldn't be a stretch to have a fully functional wall clock run off of atomic precision. Even better yet, it should have a SoC that'll hook it to your wifi network and advertise the time to anything in the area, and be accessible as part of the ntp pool.I know entire data centers that would be happy

    • I expect to see the atomic wall clock in the SkyMall catalog soon (since the Sharper Image is a shell of its former self).
  • The title of the other submission "Submission: Atomic Clock-on-a-Chip Obsoletes GPS" is inaccurate. This does not obsolete GPS at all. It makes GPS unnecessary as a precision time source in some applications, but GPS = Global Positioning System, and this doesn't provide any positioning information, it's just highly accurate and stable time source. GPS also provides that, but it provides that from multiple satellites with well defined locations. Using the time differences from 3 or more GPS satellites, you c

  • I imagine an uber-accurate clock like this could have big implications for proximity sensors that rely on time-of-flight (sonar, some lasers) and also for local, non-GPS positioning systems.
    • Good for measuring distance - there will be applications in surveying, building stress monitoring, geological measurements. Sonar doesn't go fast enough to benefit from the increased precision (Not even in water), but with this you could more easily check the placement of equipment on large construction sites via laser measurement.
  • by SnarfQuest (469614)

    Remember all those old TV shows where a group of people got together and sychronized their watches?

    Ok, set your watch .... now! ... Hold it, Jimmy is 3 picoseconds fast! We have to do it again!

  • Bought and paid for with taxes, Open Source it!

    You think it's cheap now...

  • by tttonyyy (726776) on Wednesday May 11, 2011 @07:33PM (#36101314) Homepage Journal

    I've designed kit with atomic clocks for undersea use, and specified where to procure them from. At the time (over a decade ago) rubidium clocks could be imported from Switzerland for between £1000 and £2000 with a choice of outputs (square, sine, frequency, amplitude...). They were the size of approx two Nintendo DSs on top of each other. Power consumption wasn't that bad.

    Given how long ago that was, I imagine things have improved significantly in terms of form factor and power consumption since then.

    So I'm not convinced on the headline assertions about how massive they are and the huge amount of power they draw :)

  • This will be ideal for setting up a GPS on cubesats for mars/moon. Such a system would require small amounts of circuitry since it would not be encrypted, nor taking commands. It would simply say when and where.
  • This strikes me as a bit of marketing. Everyone knows the ultra precise atomic clocks are as good as it gets, but cheap, less precise atomic clocks have been around for quite some time. In fact, you can do better with a high end temperature regulated quartz clock than a cheap atomic. And quartz offers the full range of options in between. I don't see any data on accuracy in tfa, so I'm going to assume it's slightly more accurate than a wrist watch...

    http://ieee.li/pdf/viewgraphs/precision_frequency_gene [ieee.li]

  • There's a podcast where Fossey talks about the device on the Electronic Component News website: http://www.ecnmag.com/audio/2011/01/tinkers/first-Commercially-Available-Chip-Scale-Atomic-Clock.aspx [ecnmag.com]

  • just dont take your networked widget with this thing embedded in it on a plane. chaos will ensue when you land. also avoid sudden acceleration.

"Hey Ivan, check your six." -- Sidewinder missile jacket patch, showing a Sidewinder driving up the tail of a Russian Su-27

Working...